[Stoves] Method for manufacturing moulds for making the Maputo Ceramic Stove

[Crispin Pemberton-Pigott]

UPDATE: Maputo Ceramic Stove Production

While intending to produce large numbers of stoves is a noble goal, actually getting it done can be quite a challenge.  In case you are not aware of some of the names used I will describe what things are.

While it is the intention to press the stove bodies, the tooling is not ready yet so limited production will use an older method.  It is clear that it will have to be vibrated while under pressure and that is taking time to develop.  In the meantime we are proceeding with a different system which can produce wetter moulded stoves at a high rate.  The machine used is a Jigger or 'Jigger Jolly' as it is sometimes called.  It is pictured at http://stoves.bioenergylists.org/en/mcsmould together with the mould making process.  (This text refers to that webpage.)

The Jigger is an electric machine and has a vertical shaft with a thread on the end.  The thread is visible in the picture.  Onto that shaft can be screwed any number of mould holding plates, typically with a 10 deree taper so the mould can drop in and be gripped by friction alone.  The plate is attached and moulds are dropped onto or into it.  The plate is spun, which in turn spins the mould.  Some clay is dropped into the cavity and a metal bar is pressed against the clay, forcing it against the spinning mould.  The mould and the plate can have different profiles.  In this way, a stove body (or a cup or saucer) can be formed accurately in less than a minute.  The metal plates wear out after a while.

The moulds are made from 'cast drywall' compound like Rhinocast or something specialised like ALMOD 60 which is a finely ground similar product.  They are both reasonably dimensionally stable as far as I can see.

The 'tool' spoken of here is the mould holder.  It is a steel plate which screws onto the Jigger shaft.  It has bolted to it a large nylon tapered part that the moulds sit into.  The moulds have to be larger (vertically) than the mould holder so that after the stove body is formed, the mould can easily be lifted out.  On the other hand, we have to make moulds to fit exactly into the mould holder.  The approach taken is to cast a master mould into an extended mould holder.  There is a white nylon extension ring placed on top of the mould holder, giving the extra height needed to cast the mould.

Thus the mould holder with its added ring is the mould's mould, so to speak.  In the photos a number of steps are shown whereby an inner form (made of white nylon and looking a lot like a solid stove body) is bolted to the steel plate at the base of the mould holder, creating a ring-like cavity that is the shape of the mould.  

Into this cavity the wet Rhinocast (in this case) is poured.  The following day the inner mould is removed and the mould left to dry.

The material poured into the cavity to make the moulds is not like cement which hydrates (absorbing the water chemcially).  This material literally has to dry.  As the mould holder and inner form are made of nylon, little water is evaporated so the first, or master mould, takes longer to make than subsequent copies of it.

Multiplication of the master mould

Having made a master mould, other casts are made in wooden frames to produce negatives of the outside of the master.  There can be many of these made from the master mould.  The casting material is the same gypsum-based product.  Because they are not nylon, these moulds absorb water far more quickly and have a short turn-around time.

I know the number of 'forms' 'moulds' 'inner moulds' is confusing, but bear with me.  Here is the sequence restated:

1.    The tool attached to the Jigger (the mould holder) has an extension to raise its wall high enough to cast a Master Mould.  An inner mould is needed to do this.  (Later, when producing stoves, the ring is removed.)

2.    A master mould is made and of course it perfectly fits the tool that will later spin in because it was cast inside the tool.

3.    A wooden box is made and the master mould placed in it, with the inner mould again put inside it, locating itself over a steel pin in the centre of the  wood box.  This pin is the key to accuracy.  The pin locates the inner mould which of course perfectly fits the master mould and holds it in position.  The gap between the master mould and the wooden box is filled with mix and it sets.  The separating compound used is either vaseline or Sunlight Soap.  On the nylon we are using vaseline.

4.    About one hour after the mix is cast into the box, the inner mould can be removed, as well as the master mould, leaving the box with a very fresh impression of the master mould's outside profile exactly centered on the pin.  This rapid extraction is possible because the master mould absorbs a lot of water from the mix.

5.    The master mould is left to dry for a while.  Two hours after being removed from the first box it is placed into a second box and another casting made in the same manner, producing a second mould maker.  In a day, three or four boxes can be cast from a single master mould and inner mould.

6.    After all these boxed casting are dry, the making of the moulds to produce the stove can begin.  The nylon inner mould is once again placed over the locating pin in the wooden box and bolted down.  Rhinocast is mixed and poured into the cavity.  After about an hour, the inner mould is removed and immediately re-used in the next box.  After two additional hours, the first casting can be removed from the first box because it will have dried sufficiently to be handled.  In this way, each box can produce three new moulds per shift, per day with only one nylon inner mould.  The maximum number of moulds that can be produce per shift with only one nylon inner mould is about eight which would require eight boxes.

To make a stove, the mould is dropped into the mould holder in the Jigger.  The clay is put in, the steel profile brought into position and the clay is pressed into shape.  The mould absorbs water from the clay quite quickly and the stove body can be de-moulded after about three hours.  Thus, each mould can produce two or three stove bodies per shift.

The moulds are quite heavy so the real 'work' is to move them to the Jigger, away again and to store them for a few hours in a loop of activity.  In a double shift, it is possible to run the loop five times.

Using the above process, gearing up to a production rate of 500 stoves a day from perhaps 200 moulds can be achieved in less than a month.  All the moulds will fit the Jigger properly and be very well centered so the product will be consistent.

In order to know if this process would work with the low thermal expansion PK11 clay+feldspar mixture, experiments were run to determine that PK11

a)    holds together if formed in a jigger
b)    has more self-adherence that adherence to the mould wall
c)    shrinks sufficiently to de-mould without breaking
d)    dries rapidly enough to keep the number of moulds required to an affordable level
e)    is not filled with small fractures when drying, leading to early failure

We found that PK11 dries quite rapidly compared with regular clay (about twice as fast) and shrinks less with a Jigger-compatible water content.  This is probably because the clay content is only 40%.  It is strong enough to handle and carve.  It takes glaze well and is extremely strong after firing.  The charcoal has to be removed (by firing) before glaze can be put on.  This means a glazed stove has to be fired twice.

In these photos, only the process of making the master mould is shown.  

END

C Pemberton-Pigott
18 Feb 2007